Key Points
- •
Otitis media remains an extremely common disorder; vaccination strategies are decreasing its incidence.
- •
Recent guidelines highlight the need for more careful diagnostic criteria.
- •
Initial treatment can include either initiation of antibiotics or a period of observation, depending on patient age and severity of presentation.
- •
Frequent otitis media is a key feature of primary immune deficiency.
- •
Respiratory allergy contributes to chronic otitis media with effusion (OME), but more study is needed to fully define the impact of allergy on this common condition.
Introduction
Acute otitis media (AOM), and the related otitis media with effusion (OME), are the most common diseases requiring pediatric care in the first decade of life, except for viral upper respiratory infections. The costs of primary and specialty care, as well as the indirect costs incurred by the family, are enormous. In 2013, estimates for the direct cost of treating otitis media (OM) in the USA totaled $ 2.88 billion, not including indirect costs. Indirect costs incurred by families (such as hours of work lost, transportation, etc.) may equal the direct costs, doubling the financial burden. There have been significant advances in the past 40 years in understanding of the pathogenesis, pathophysiology and immunopathology of OM, leading to a decrease in physician visits and antibiotic prescriptions for this illness. Improved recognition, a willingness to observe the evolution of less severe cases and widespread pneumococcal vaccination likely have contributed to this improvement. This chapter provides a review of the epidemiology, pathogenesis, Eustachian tube (ET) physiology and immunology of OM as well as medical and surgical therapies employed to treat it. It also provides information on the potential role of allergy in the pathogenesis of this prevalent condition.
Definitions
OM is characterized by acute or chronic inflammation of the middle ear (ME). AOM is typically preceded by or associated with viral upper respiratory tract infections (URTI); up to 37% of viral URTIs may be complicated by OM. Persistent OM is defined as persistence of symptoms and signs of ME infection despite antimicrobial therapy (i.e. treatment failure) and/or a relapse of AOM within 1 month of completion of antibiotic therapy. When two episodes of OM occur within 1 month, it may be difficult to distinguish recurrence of AOM from persistent otitis media (relapse). Recurrent AOM is defined as having 3 or more episodes of AOM in 6 months or 4 episodes in 12 months ( Box 25-1 ).
Acute otitis media
Recurrent acute otitis media
Persistent acute otitis media
Otitis media with effusion
Chronic otitis media with effusion
AOM often evolves into OME, chronic middle ear effusion (MEE) without signs or symptoms of acute infection. After an episode of AOM, 60% to 70% of children have OME at 2 weeks, decreasing to 40% at 1 month and 10% to 25% at 3 months. Chronic OME is defined as OM lasting for 12 weeks . Children with certain sensory, physical, cognitive and behavioral conditions are particularly vulnerable to the hearing loss, speech, language and learning problems associated with OME.
Epidemiology
The peak incidence for AOM is during the first 2 years of life, and most initial cases occur between 6 and 12 months of age. By 1 year of age almost 50% of children have had at least one episode of AOM, and two thirds by 3 years. Three or more episodes of OM occur in 10% of children by 1 year of age and in 33% by 3 years of age. Tos and colleagues reported a point prevalence of OME of 13% during the first 2 years of life. In 4- to 6-year-old children, point prevalence decreased to 7% and then further decreased to 2% to 4% in 8- to 10-year-old children. Whereas population-based studies in the USA and Finland suggested that OM was increasing toward the end of the 20th century, changes in healthcare systems, access to care, patterns of using services and awareness of OM may be partially responsible for these increases. The widespread use of pneumococcal vaccination appears to play a positive role in reducing the overall incidence of acute OM and subsequent complications, although the effect appears to be restricted to early infancy.
Multiple factors increase the risk of OM in children. The best-defined risk factor in the development of AOM is a preceding viral URTI. A prospective cohort study by Wald and colleagues reported that 25% to 40% of URTIs in children from birth to 3 years of age were accompanied by an episode of AOM, most commonly under 1 year of age.
Universally, males are affected more than females. Indigenous populations, such as North American Indians, native Canadians and Polynesian children, have a much higher incidence than white children. Compared with bottle-feeding, breastfeeding for at least 6 months is associated with a decreased risk of acute otitis or recurrent otitis during the first year of life. This protective effect on both the frequency of URTIs and the resultant AOM may not be seen with shorter durations of breastfeeding. Cigarette smoking by the parents, especially the mother, is a significant risk factor for AOM during the first year of life. Children whose parents or siblings have had a history of chronic otitis have a higher incidence than those with no family history. A cohort study of 2,512 children in Finland concluded that, while family size, low socioeconomic status and cigarette smoking were all individually correlated with an increased risk of recurrent AOM, they were all interdependent variables. The association of these factors with AOM was best accounted for by the strong correlation with attendance at large (≥ 20 children) daycare centers as well as short duration of breastfeeding. A study of 175 sets of twins and triplets followed from birth indicated a genetic component to susceptibility to AOM. The estimate of discordance of AOM in monozygotic twins was 0.04 compared with 0.49 in dizygotic twins ( P < .005). Numerous studies confirm an augmented risk of AOM, recurrent otitis media (ROM) and OME as the number of children in the childcare setting increases. Other factors such as preterm birth and greater number of siblings in the household also increased a child’s risk of AOM. Specific conditions such as Down’s syndrome, craniofacial anomalies, ciliary dyskinesia syndromes and primary and secondary immune deficiency syndromes are associated with increased risks of AOM ( Box 25-2 ).
Environmental Factors
Viral upper respiratory tract infection
Daycare attendance
Cigarette smoke (passive)
Low socioeconomic status
Environmental pollution
Host Factors
Male gender
Genetic predisposition
Premature birth
Not breastfed
Supine bottle-feeding
Immune deficiency (primary and secondary)
Craniofacial abnormalities
Eustachian tube dysfunction
Cilia dysfunction
Allergic rhinitis
Using tympanometry, Mandel and Casselbrant found that asymptomatic OME is relatively frequent in daycare settings, especially in the winter months. Repeated evaluations indicated that many cases resolved spontaneously without therapy and effusions could persist for up to 6 months without overt symptoms.
Whether allergy predisposes children to AOM and OME is an area of controversy. Scandinavian and US studies have shown that, during the first 3 years of life, there is no association between AOM and allergic disease. Allergy is cited as a potential risk factor for OME, especially in children needing surgical intervention. Early clinical studies suffered from significant methodologic limitations such as lack of control groups or unclear definitions of allergy. For example, one study of children with chronic OM referred for placement of ventilation tubes found that approximately half the children had positive allergy skin tests or increased serum immunoglobulin E (IgE) antibodies to specific allergens. However, there was no comparison to normal controls, nor was it clear whether these children had clinical allergies or asymptomatic sensitization. More recent studies have used prospective birth cohorts to examine the risk of OME in children with allergic disease. The COPASC cohort from Copenhagen found little effect of allergic rhinitis on OME in early life, but a strong correlation at age 6. This was not only due to obstruction, but potentially to effects of Th2 inflammation. The LISA cohort from Germany found similar associations at age 6. They also examined if early OM predicted onset of allergic disease; they found positive correlations for late-onset atopic eczema and asthma, but not for allergic rhinitis. Several in vitro studies substantiate a pathophysiologic link between allergy and OME.
No association between OM and food ingestion has been found. A large unblinded trial showed some efficacy of elimination diets in OME. One study suggested a link between IgE-mediated cow’s milk allergy and recurrent otitis media, however the effect could be completely accounted for by the presence of respiratory allergies in these children. There has been little progress and no significant peer-reviewed studies in this area in the past decade.
Pathophysiology
Structure and Function
Otitis media is a disease of the upper respiratory tract. Ventilation of the ME is accomplished via the ET from the posterior nasopharynx. Middle ear effusions in children are most often related to abnormal ET function. The ET provides an anatomic communication between the nasopharynx and the ME. Like mucosa elsewhere in the respiratory tract, the ET lining contains mucus-producing cells, ciliated cells, plasma cells and mast cells. Unlike the bronchial tree, the ET is usually collapsed and thus closed to the nasopharynx and its contents. The ET, like the bronchi, serves several physiologic functions. It protects the ME from nasopharyngeal secretions, drains secretions produced within the ME into the nasopharynx, ventilates the ME to equilibrate pressures and replenishes oxygen in the ME. In normal tubal function, intermittent opening of the ET maintains near-ambient pressure in the ME cavity. It is suspected that in cases in which active swallowing is inadequate to overcome tubal resistance, the tube remains persistently collapsed, resulting in progressively negative ME pressure. This abnormal pressure appears to be common in children. Periodic or persistently high negative pressure may be pathologic and associated with abnormal ET function and may lead to AOM.
Eustachian Tube Obstruction
Two types of ET obstruction, mechanical and functional, could result in acute or chronic OME ( Box 25-3 ). Intrinsic mechanical obstruction may result from inflammation of infection or allergy, whereas extrinsic obstruction may result from enlarged adenoids or, in rare instances, nasopharyngeal tumors. Experimentally, allergic rhinitis provoked in patients with a history of allergy has been associated with the development of ET obstruction. A persistent collapse of the ET during swallowing may result in functional obstruction, which appears to be related to increased tubal compliance, an inefficient, active opening mechanism by the tensor veli palatine muscle, or both. The angulation of the craniofacial base changes with age, improving the tensor veli palatine muscle after puberty. Additionally, in infants and younger children the cartilaginous support of the ET is less robust.
Mechanical obstruction
Intrinsic
Infectious inflammation
Allergic inflammation
Extrinsic (peritubular)
Adenoidal hypertrophy
Nasopharyngeal tumor
Functional obstruction
Poor tensor veli palatini muscle function
Increased tubal compliance
Pathogenesis
A role for ET dysfunction in the pathogenesis of AOM during a viral URTI is supported by multiple clinical and experimental studies. Studies reported tubal dysfunction in children and adults with natural viral URTI, experimental infection and animal models.
Rhinovirus infection results in significant increases in nasal inflammation, impaired tubal function and abnormal ME pressures in more than 40% of subjects, and asymptomatic OM in approximately 2% of subjects. These events occurred sequentially and in descending frequency, supporting a causal pathway. This pattern occurred in infection with rhinovirus, influenza A virus, and coxsackievirus A and influenza. In the majority, OM was asymptomatic with experimental infection and the recovered effusion was negative by culture for viruses and bacteria but positive for influenza A and Streptococcus pneumoniae by polymerase chain reaction (PCR). The inflammatory response to acute infection is a key part of the pathophysiology of the disease. Using a murine model of ME and H. influenzae infection, Ryan and colleagues documented that ME mucosa rapidly undergoes hypertrophy and within 24 hours exhibits edema, mucosal thickening and lymphocytic infiltrate. This progressed over a 5- to 7-day period and resolved. Similar pathology was noted in larger animal models of AOM, including rats, guinea pigs and chinchillas. Middle ear effusions were accompanied by inflammatory cytokines, including TNF-α, IL-1, IL-8, IL-10 and IL-6, which diminished over 48 to 72 hours. The inflammatory infiltrate was potentiated by chemokines produced in the infectious response. What is lacking in the animal models of acute OM are experiments mimicking both early viral infection and subsequent bacterial superinfection characteristic in humans. Most models induce OM via introduction of bacteria to the ME via the tympanic membrane (TM) or the bulla (i.e. the ME in mice). While this duplicates the pathogens found within ME fluid in AOM, it does not completely mimic AOM pathogenesis. A small number of models have employed co-exposure to viruses and bacteria, primarily to assess therapeutic approaches. A novel, potentially highly important contribution of animal models is the discovery of genes that may predispose to OM. Genetic work has identified the key role of innate immunity in the inflammatory response of the ME. It is well recognized that, in addition to humoral immune defects, subtle defects in mucosal immunity such as mannose binding ligand defects render children more susceptible to OM. Using murine models, a crucial role for proteins in the Toll receptor pathways was identified. MyD-88 signaling may also be crucial for defense against bacterial pathogens in AOM. Indeed, defects in chemokines such as IL-8, MCP-1 and CCL3, which play integral roles in promoting early neutrophilic inflammation, can predispose animals to AOM. In contrast, deleting CCR5 can ameliorate the severity of bacterial OM.
The role of mast cells and innate immune responses in AOM has been the subject of multiple studies. Mast cells, as primarily mucosal leukocytes, are the most common hematopoietic cells found in the normal ME, including the lining of the ME, ET and the TM. This has led to a focus on allergy in the pathogenesis of otitis. However, it is unlikely that the primary role of the mast cell in OM is via its ability to bind IgE on FcεR1, but rather its role in binding IgG via Fc gamma receptors. In cKit knockout mice, which are unable to produce mast cells, infection induced into the ME did not cause inflammation, mucosal changes or remodeling found in mast cell sufficient mice or cKit knockout mice that had received infusions of normal mast cells. ME fluid harbors mast cell mediators in both acute and chronic models of OM, underlining the role of mast cells in the normal host response against bacteria.
Another major advantage of the expanding role of murine models is the ability to detect new genes that may predispose to OM. Defects that affect the development of ET and ME include mutations in the transcription factor EVI1 and the eyes absent homolog Eya1 . A mouse model of deafness known to develop chronic suppurative OM, Jeff, carries a mutation in an F-box gene, Fbxo11 , predisposing to development of cleft palate. Even well-known genetic syndromes such as DiGeorge/velocardiofacial syndrome have been advanced in the study of OM pathogenesis by murine models. This is complemented by genome wide association studies in several cohorts of human subjects that identified predisposing single nucleotide polymorphisms in innate immune pathways such as IL-1β and CXC3R1.
Etiology
Acute Otitis Media and Otitis Media with Effusion
Bacteria are found in approximately 60% to 70% of children with AOM who undergo tympanocentesis. Historically, Streptococcus pneumoniae , Moraxella catarrhalis and nontypable H. influenzae were the predominant causative bacteria in AOM. Group A β-hemolytic Streptococcus, Staphylococcus aureus and, more rarely, anaerobes, account for a minority of cases. Viruses alone are recovered from ME fluid in about 15% of cases. Since the routine institution of the heptavalent pneumococcal conjugate vaccine (PCV7), the microbiology of AOM has changed ( Table 25-1 ). About 50% of children requiring myringotomy for AOM refractory to second-line antimicrobial treatment had positive cultures, most growing organisms such as S. pneumoniae , Staph. aureus and coagulase negative Staphylococcus . Finnish and US trials have demonstrated that PCV7 has resulted in a 6% overall reduction in the clinical incidence of AOM which may diminish ROM and the need for tympanostomy tube placement. The majority of the effect is in younger children, while children who were vaccinated later in life had less effect from PC7. Studies with more extensive vaccine coverage will be highly instructive in determining if there is a broader effect; in terms of healthcare dollars, 5% to 6% difference has significant impact.
| Bacteria | Percentage |
|---|---|
| Haemophilus influenza e | 51 |
| Streptococcus pneumoniae | 38 |
| Moraxella catarrhalis | 12 |
| Staphylococcus aureus | 5 |
| Two pathogens | 12 |
| No growth | 30 |
Penicillin-resistant S. pneumoniae is a significant clinical problem and is found in up to two thirds of pneumococcal isolates. Approximately 50% of H. influenzae and 90% of M. catarrhalis strains produce β-lactamase, making them resistant to amoxicillin. Additionally, while antibiotic prescriptions for AOM have decreased somewhat, there is an increase in treatment of OM with broad-spectrum antibiotics including single-dose third-generation cephalosporins. This may well impact antibiotic resistance patterns.
Previously it had been assumed, incorrectly, that chronic middle ear fluid (MEF) effusions were sterile. In several studies, about 50% of the chronic, persistent ME effusions had positive cultures for bacteria whose microbiology was similar to that found in acute otitis.
Mediators of Allergy and Otitis Media with Effusion
Studies have focussed on analysis of the inflammatory infiltrate found in MEEs from OME. There is no uniform infiltrate; both Th1 and Th2 inflammation have been found in analyses of ME fluid. MEEs with infiltrates characteristic of allergic inflammation have been studied in detail. Eosinophilia and proteins derived from eosinophil degranulation are unique features of ME disease in allergic patients. Statistically significant differences in major basic protein and IL-5 mRNA were found in ME biopsy specimens, suggesting both eosinophil recruitment and degranulation in the ME. Elevated levels of IL-4, mast cell-derived tryptase, eosinophilic cationic protein and RANTES (regulated upon activation normal T cell-expressed and secreted) were all found in higher concentration in children with atopic backgrounds compared with nonatopic children.
Using a cohort of 75 children skin tested prior to surgery for OME, Sobol and colleagues and Nguyen and colleagues studied cellular components and cytokine expression in MEEs of patients undergoing tympanostomy tube insertion. The incidence of atopy (positive skin tests to at least one of 12 common allergens) in these three studies was 24% to 30%. Atopic children had significantly higher levels of eosinophils, T lymphocytes, and IL-4 + and IL-5 + cells on immunohistochemistry compared with the nonatopic group, and a trend toward higher mast cells and basophils. Nonatopics had higher IFN-γ + cells. Th2 cells and cytokines were found in ME fluid in atopic children and in biopsy specimens from adenoid tissue and the torus tubarius, demonstrating a strong correlation between allergic inflammation in MEE and the upper airway. A UK study with an incidence of atopy of 7% defined four groups of children with OME based on their MEE infiltrate and cytokine profiles. Two groups were predominantly Th1 (subacute and chronic), one had Th1-Th2 overlap, and one was strongly Th2. A strong correlation between mucin production and the Th2 cytokines IL-4 and IL-13 was observed.
Diagnosis of Otitis Media
Acute Otitis Media
The American Academy of Pediatrics (AAP) and American Academy of Family Physicians released clinical practice guidelines in 2013 outlining the diagnosis of AOM. The guidelines stipulate three criteria that must be fulfilled. There must be acute, abrupt onset of signs and symptoms of AOM such as otalgia, otorrhea, irritability and fever. There also must be a documented MEE. One can document this on examination of the ME by noting one of the following: a bulging TM, an air-fluid level behind the TM, otorrhea, or limited TM mobility on tympanometry, pneumatic otoscopy or acoustic reflectometry. The patient must also have signs or symptoms of inflammation in the ME, such as distinct erythema of the TM. However, crying and/or fever can both result in an erythematous TM. The diagnosis of AOM is often difficult in infants who are too young to clearly express themselves. They often have co-existing viral URTI, and it is often a challenge to clear the external ear canal of cerumen. The ultimate management of the infant or child will differ depending on the physician’s degree of certainty of the diagnosis of AOM. The most recent guidelines also differentiate between children who are under 24 months and older children and classify AOM as mild, moderate or severe, based on visualization of the tympanic membrane and pneumatic otoscopy. These stricter criteria may assist in decreasing unnecessary antibiotic prescriptions.
Chronic Otitis Media with Effusion
The AAP, American Academy of Family Physicians, and American Academy of Otolaryngology and Head and Neck Surgery developed clinical practice guidelines describing the diagnosis and management of OME in children. According to these evidence-based guidelines, during follow-up of all children with OME, it is important to document the laterality, duration of effusion and presence and severity of any associated symptoms at each clinical assessment. The presence of OME can be confirmed by a combination of visual inspection, tympanometry and pneumatic otoscopy. Children at ‘high risk’ for development of speech, language or learning problems as a result of MEE causing hearing loss should be promptly evaluated and may need more timely surgical intervention than low-risk children. A low-risk child with OME can be managed with watchful waiting for 3 months from the onset of the effusion (if known), or from the date of diagnosis. All children with OME lasting > 3 months should have a hearing test and be reexamined at 3- to 6-month intervals until the effusion is no longer present, significant hearing loss is identified or structural abnormalities of the eardrum or ME are suspected ( Figure 25-1 ).
Chronic Otitis Media and Allergy
Some children presenting with OM have associated rhinitis. If chronic, it is important to determine whether this rhinitis is infectious or allergic. Prolonged, perennial or recurrent seasonal rhinitis with itching and sneezing suggests an allergic basis, as does co-existent allergic conjunctivitis. A family history of allergy and/or a personal history of atopic dermatitis, allergic asthma and food allergies also raise clinical suspicion of allergic rhinitis.
It is advisable that children with persistent OME be screened for allergic rhinitis by taking a clinical history addressing symptoms, signs and timing for allergic rhinitis. If this is suspected, these children should be referred to a specialist in allergy for further evaluation and investigations. Skin prick testing is preferred to serologic anti-IgE antibody tests for the detection of IgE antibodies to specific inhalant allergens because of the increased clinical sensitivity and lower cost. For either test result to be considered clinically relevant there must be a correlation between exposure to a particular allergen and clinical symptoms. Total serum IgE levels are not useful, as they do not define specific allergen sensitivities.
Diagnostic Techniques
Physical Findings
Otoscopic inspection requires visualization of the TM. The normal TM is thin, translucent, neutrally positioned and mobile. The bony ossicles, particularly the malleus, are generally visible through the TM. Adequate assessment requires that the physician take note of the TM’s thickness, degree of translucency, position and its mobility to applied pressure. A bulging eardrum and air bubbles or air-fluid levels indicate the presence of excessive ME fluid and document effusion. Various degrees of bulging may assist in better classifying the severity of the illness, whereas hyperemia and erythema alone without fluid or indications of pressure changes are much less sensitive indicators. The ear canal may be filled with pus which, when removed, will usually reveal an inflamed TM with perforation ( Figure 25-2 ).
